Q3 Cheat Sheet Bivariate Frequency Distribution

Cheat Sheet (Bivariate Frequency Distributions)

(1) a. N = , K = 2 * 2, (K1 = 2, K2 = 2) Where N is the number in the sample, K1 is the nominal Dependent Variable, and K2 is the nominal Independent Variable. Both variables have exactly two levels.

b. Ascertain available cell frequencies from the information given

(2) Set up a table with 6 rows and 4 columns.

(3) In “Row 1” place the labels as shown for the independent variable

(Column 1) (Column 2) (Column 3) (Column 4)

(Row 1) Labels label level 1 label level 2 total

(Row 2)

(Row 3)

(Row 4)

(Row 5)

(Row 6)

(4) In “Column 1” place the labels as shown for the dependent variable

(Column 1) (Column 2) (Column 3) (Column 4)

(Row 1) Labels label level 1 label level 2 total

Row 2) Count level 1

(Row 3) Percent level 1

(Row 4) Count level 2

(Row 5) Percent level 2

(Row 6) Total

(5a) In columns 2 and 3, insert the available frequencies for those “cells.” Some Cell frequencies must be calculated.

(Column 1) (Column 2) (Column 3) (Column 4)

(Row 1) Labels label level 1 label level 2 total

(Row 2) Count level 1 XX XX XX

(Row 3) Percent level 1

(Row 4) Count level 2 XX XX XX

(Row 5) Percent level 2

(Row 6) Total XX XX XX

(5b) Place the row totals and the column totals (marginals) in parentheses.

(Column 1) (Column 2) (Column 3) (Column 4)

(Row 1) Labels Label cat 1 label cat 2 total

(Row 2) Freq cat 1 XX XX (XX)

(Row 3) Percent cat 1

(Row 4) Freq cat 2 XX XX (XX)

(Row 5) Percent cat 2

(Row 6) Total (XX) (XX) (XX)

(6a) Determine the cell percentages: cell% = cell Frequency / column total

(Column 1) (Column 2) (Column 3) (Column 4)

(Row 1) Labels label level 1 label level 2 total

(Row 2) Freq level 1 XX XX (XX)

(Row 3) Percent level 1 XX% XX%

(Row 4) Freq level 2 XX XX (XX)

(Row 5) Percent level 2 XX% XX%

(Row 6) Total (XX) (XX) (XX)

(6b) Circle row 3

(7) Identify the minimum value in Column 4, the smallest number in Column 4 is E1.

(8) Identify the lowest value in Column 2, and add it to the lowest value in Column 3. This produces E2.

(9) Compute Lambda, the Magnitude of Effect (MOE). λ = (E1 - E2) / E1

Lambda is the proportional reduction in error. .20 is weak, .40 is moderate, .60 is strong.

There is no effect (MOE) if there is not a statistically significant relationship between the two variables. That relationship is typically established using Chi-Square procedure, which is not discussed until later in the semester. Therefore, interpret Lambda with caution.